exploded battery

[philSweet]

The trick is to minimize paralleling of batteries. Most are just not inherently fail-safe if paralleled. RVs and boats and diesel pickups often parallel two batteries without fuses. In separate enclosures or in an open bay, this does not seem to lead to fires during a short using deep-cycle batteries or even start batteries in pickups. If you find yourself wanting to parallel more than two, get bigger batteries. Or shift to higher voltages. The cost of fuses and wiring for a many-in-parallel scheme that actually will do the job reliably and not cause other bothers in a typical marine installation would be much greater than just going to a different configuration.

And yes, you typically have to replace all the batteries in a bank unless you are standing there when the fault occurs, so I tend to overwork a smaller bank and replace it more often. I'd rather replace two batteries every two years than three every three years, which is about what happens running the banks at equal power draw cycles. This is another way to reduce paralleling.

Another issue with parallel strings is charging. You are less likely to create an over-current condition in a 2 in parallel configuration than in a many-in parallel configuration with the charger when one cell dies. Modern chargers can detect bad batteries and shut down, but not if the batteries are in parallel.

from rowerwet:

The cells were being torn apart by magnetic fields induced by high current.

Which is why you need to destructively test complete installations, not just components, because geometry matters. Imagine for a moment about 20 sub batteries arranged in a ring around a missile in a silo. Then a short. Current rise on the order of 1 billion amp/sec. Peak current on the order of 1 million amps going in a circle around the missile. That's quite a decent inductor, with a 100,000 pounds of rocket fuel that is ignited electrically sitting in the middle. The mechanical stresses are pretty substantial as well. It happened, once. Fortunately, the place is designed to take an EMP pulse of similar magnitude. The configuration was redesigned after that.

 

[TerryKing]

Oh good! Something I know about

I use/sell thousands of DS18B20 electronic temperature sensors. See:

http://arduino-info.wikispaces.com/Brick-Temperature-DS18B20

Scroll down a ways to the waterproof sensors in stainless steel tubes.

I have these in 1M, 4M and 10M length cables; It's nice to NOT have to splice stuff in many cases.

Evey sensor of this type has it's own unique address. So you can parallel (not a Bad Word in this case) several sensors and run them on a single Microcomputer I/O pin. See the example here with 4 sensors:

http://arduino-info.wikispaces.com/MultipleTemperatureSensorsToLCD

Once you have the data you need to decide what to do with it. In most 'alarm' situations a good system will calculate the rate of change of temperature as well as check against a maximum value. Any fast change means trouble.

The same system could be showing you water temperature outside the hull, cooling system temperatures, cabin temperature, water heater temperature etc etc.

A variety of automatic systems can be made if you learn about this stuff. And it's cheap these days. See:
http://ArduinoInfo.Info


DISCLAIMER: Mentioning stuff from my own shop... but you can find things on Ebay...

Maybe start here?
http://yourduino.com/sunshop2/index.php?l=product_detail&p=441

CDK if you want to dive into this stuff email me; I'll send you one of these..

 

[TeddyDiver]

The trick is to minimize paralleling of batteries. Most are just not inherently fail-safe if paralleled. RVs and boats and diesel pickups often parallel two batteries without fuses. In separate enclosures or in an open bay, this does not seem to lead to fires during a short using deep-cycle batteries or even start batteries in pickups. If you find yourself wanting to parallel more than two, get bigger batteries. Or shift to higher voltages. The cost of fuses and wiring for a many-in-parallel scheme that actually will do the job reliably and not cause other bothers in a typical marine installation would be much greater than just going to a different configuration.
.........
Another issue with parallel strings is charging. You are less likely to create an over-current condition in a 2 in parallel configuration than in a many-in parallel configuration with the charger when one cell dies. Modern chargers can detect bad batteries and shut down, but not if the batteries are in parallel.

Big thumb

 

[Angélique]

There ARE Normally-closed solenoid type relay. But the system would have to have power from the "other" batteries to disconnect a battery. And keep it disconnected until manual intervention.

There must be a relay with actuation like the usual momentary car door lock mechanism... Hmmm...

I was thinking on each battery a maybe specially designed solenoid switch where the connect and break positions are both spring retained (like the light switch at home) and the switch is pulled over a dead point to switch to the other position. When the monitoring system calls for it then the solenoid is temporary activated to pull the switch over the dead point to the break position. After the problem (read out the monitoring system for it) that caused the break is solved then the switch is manually pulled over to the connect position again, or a second temporary activated solenoid is installed on the switch to do the reset.

If such a solenoid switch isn't on the market then specially designed shouldn't be a problem if the demand for it is big enough, just like the ‘Distributed Power Systems / Digital Power Distribution’ (that should operate the solenoids) recently didn't exist and now there seems to be a developing demand for it.

A separate battery for power supply of the monitoring system and to operate the solenoids, while the main system is guarding the monitoring system and also is able to activate the solenoids when needed, will enhance the safety level.

 

[Barry]

The exploded battery in the initial post was reported to be part of a battery bank, apparently with batteries wired in parallel. In that situation an internal short in one battery involving one or two cells lowers the potential across that battery compared to the other batteries in parallel, and very large currents result, particularly through the battery with the short, as noted by CDK, the OP. The un-shorted cells in the battery with short are being charged at a very rapid rate by the other batteries in parallel. Hydrogen is produced at a rapid rate and the battery can become extremely hot. The cables connecting the batteries may also overheat. An explosion may follow. However if each battery has individual overcurrent protection such as a fuse, fusable link or other device which opens and isolates the shorted battery then the battery with the short will be protected and the battery bank will continue to function without the battery with the short.

Batteries in series do not have this problem. The current is the same through all batteries, even if one battery develops an internal short. A single overcurrent protection on batteries in series is presumably sufficient.
?

"Batteries in series do not have this problem."

Not sure why they would not have the same problem.
Lets say we wire 2 6 volt batteries in series and one cell is shorted. The short is not to ground, but rather the short is internal, from one side of the cell to the other. Instead of having 6 each 2 volt cells ( more than 2 V but lets keep it simple) making up the 12 volt, we now have 5, so the battery voltage now becomes a 10 volt battery being charged by a 12 volt charger. This charger/rectified alternator will continue to supply current to the 10 volt battery, as it senses a low voltage situation. The plates will become over charged, heat will increase, hydrogen will be produced and we are back to the same issue.

?????

 

[TerryKing]

I was thinking on each battery a maybe specially designed solenoid switch where the connect and break positions are both spring retained (like the light switch at home) and the switch is pulled over a dead point to switch to the other position. When the monitoring system calls for it then the solenoid is temporary activated to pull the switch over the dead point to the break position. After the problem (read out the monitoring system for it) that caused the break is solved then the switch is manually pulled over to the connect position again, or a second temporary activated solenoid is installed on the switch to do the reset.

If such a solenoid switch isn't on the market then specially designed shouldn't be a problem if the demand for it is big enough, just like the ‘Distributed Power Systems / Digital Power Distribution’ (that should operate the solenoids) recently didn't exist and now there seems to be a developing demand for it.

A separate battery for power supply of the monitoring system and to operate the solenoids, while the main system is guarding the monitoring system and also is able to activate the solenoids when needed, will enhance the safety level.

Angélique this is a good idea.

Back in a previous century I was a Broadcast Engineer. Some big transmitters had a current overload relay that would operate only when some short circuit or arc over happened. That relay mechanically latched OFF and had to be reset manually by a human pushing a button.

A battery disconnect relay like that would be good; it would only be tripped to disconnect the battery on a overcurrent or overtemperature fault.

AND being able to easily isolate batteries and test them would be a big advantage I think.

Now to find such a beast

 

[DCockey]

"Batteries in series do not have this problem."

Not sure why they would not have the same problem.
Lets say we wire 2 6 volt batteries in series and one cell is shorted. The short is not to ground, but rather the short is internal, from one side of the cell to the other. Instead of having 6 each 2 volt cells ( more than 2 V but lets keep it simple) making up the 12 volt, we now have 5, so the battery voltage now becomes a 10 volt battery being charged by a 12 volt charger. This charger/rectified alternator will continue to supply current to the 10 volt battery, as it senses a low voltage situation. The plates will become over charged, heat will increase, hydrogen will be produced and we are back to the same issue.

?????

The major and critical difference between the scenario you describe involving a short in one cell of a battery in series and a short in cell in a battery in parallel is the very large difference in the amount of current trying to charge the battery with the shorted cell.

Let's start with the battery in series. For this example it doesn't matter if the physical construction has one case with 6 cells in series (a "12 volt" battery) or two cases with 3 cells (two "6 volt" batteries) in series. If one cell shorts the voltage will be reduced but the current will be limited by the design and capabilities of the charging system, whether it is an alternator driven by an engine or a charger powered by a 120V or 240V circuit. The charging rate will be the same as if the cells had equally discharged so that the total voltage was 10 volts, and the battery should be able to deal with that charging rate without great drama. ( For a typical automotive type system the charging rate at 10 volt battery voltage will probably be on the order of 20 amps or so.) Left charging long enough and the water will probably be boiled from the cells but that is not particularly dramatic.

Now consider the situation with several batteries in parallel and one cell in one battery is shorted so that one 10 volt battery is in parallel to one or more 12 volt batteries. The current flowing into the battery with the shorted cell will only be limited by the resistance of the cables and the internal resistance of the batteries. That current will be hundreds or even thousands of amps for a typical system, particularly if three or more batteries are in parallel. The effect of that extremely large current trying to charge the battery with the shorted cell will be significantly greater than the effects of normal charging current, and is likely to have catastrophic consequences.

 

[Barry]

DC
so your view is that with one battery at 10 volts and the others at 12 would cause current inputs into the 10 volt battery in the thousands of amps???

If we ignore battery charger issues for now
Then if I was to put say a 150 amp slow burn fuse or circuit breaker between the positive buss bar and each positive post, then this could solve future problems. I would be able to pull up to 150 amps off each battery, much more than needed but if a cell goes and the amps go up to this dead cell shorted battery, it would drop off the grid so to speak.

I would like to see any calculations of current draw of one dead cell 12 volt battery with say a couple of 800 ah batteries wired in parallel if any one can do this.

 

[DCockey]

DC
so your view is that with one battery at 10 volts and the others at 12 would cause current inputs into the 10 volt battery in the thousands of amps???

Magnitude of the current depends on the number of batteries and their characteristics as well as the resistance of the wires connecting the batteries but thousands of amps would be possible with some systems used in boats.

If we ignore battery charger issues for now
Then if I was to put say a 150 amp slow burn fuse or circuit breaker between the positive buss bar and each positive post, then this could solve future problems. I would be able to pull up to 150 amps off each battery, much more than needed but if a cell goes and the amps go up to this dead cell shorted battery, it would drop off the grid so to speak.

That's the idea of a fuse or fusible link for each battery.

 

[Barry]

DC
I was hoping to find out the current draw to a 12 volt battery that had one dead cell but hooked to several other batteries would be to see how big of fuse that would be required.
The fuse would still have to be big enough to serve the equipment draw at max but small enough to trip if a cell went dead.

Re the charging system
When we took delivery of our boat, it was a few years old but had about 150 hours, the seller had a remote battery charger hooked to one engine battery. I did not give this much thought as there were multiple systems to get acquainted with.
(much later we discovered that the start batteries were not connected to a charge system other than the alternators on the engines, ie no shore power charger)
As part of the orientation, the fellow took off the top of the sea strainer and on the Volvo d-6's, there are at the top of the engine.
12 hours into the trip, an engine compartment alarm went off, the fire suppression did not trip, so I opened the hatch to a water vapor cloud. Evidently the fellow did not tighten the strainer lid enough which caused a leak which in time tripped the alarm.
While I was crawling around the compartment I noticed that one battery had overcharged, bulged and looked like it would burst. Dead cell, no other batteries in series or parallel. So it took about 12 hours of running time, with only the alternator charge output, which might be 90 amps, to almost explode this battery.

And this is the dilemma. If I put in say 150 amp breakers, on the start batteries, this could repeat itself, as well as the same 150 amp breakers on the house bank 5500 ah capacity, the charging system and or the 4 batteries feeding the 5th dead cell battery could still cook a few batteries with a possibility of a hydrogen explosion

 

[DCockey]

If you have a five batteries in parallel, and a cell in one battery shorts, the current from the other batteries trying to charge the battery with the shorted cell will be many times the current from the charging system. Something dramatic will probably happen quickly, probably seconds or minutes rather than hours, unless the battery with the bad cell is isolated. Once the bad battery is isolated it will not be charged.

Barry, it sounds like you are concerned about a single battery being overcharged. An overcurrent fuse won't do anything to prevent over charging.

 

[Angélique]

Now to find such a beast

Good Luck, and put them into your sales program . .

 

[Angélique]

BTW, an automatic break system on the battery should be mandatory on cars after the airbags has gone off, cars have enough electronics to operate the system. It could prevent a short circuit and fires after an accident, and can guard the battery along the way . . :idea:

 

[Barry]

If you have a five batteries in parallel, and a cell in one battery shorts, the current from the other batteries trying to charge the battery with the shorted cell will be many times the current from the charging system. Something dramatic will probably happen quickly, probably seconds or minutes rather than hours, unless the battery with the bad cell is isolated. Once the bad battery is isolated it will not be charged.

Barry, it sounds like you are concerned about a single battery being overcharged. An overcurrent fuse won't do anything to prevent over charging.

I have both situations in our boat. Each engine has a single battery isolated from any other battery and the ability to be charged from only the alternator on the engine.


Our house bank is 5 1100 ah agm batteries in parallel. When one cell went, I had to replace all 5 batteries due to the heat issue

The focal point is the size of the fuse/circuit breaker to isolate each battery in case on battery develops a shorted cell. Each battery fuse must be able to carry the current to feed the requirements (in my case one fifth) of the equipment load.
Just to put a number on it for discussion, say each battery has to supply 150 amps.
If the current draw from the 4 good batteries into the 5th dead cell battery is more than 150 amps, the breaker/fuse will trip.

If the current draw from the 4 good batteries into the 5th dead cell battery is less than 150 amps, say 100 amps, then I will still have the same overcharging conditions

My OPINION ONLY at this point, is that a one dead cell battery with a voltage at 10volts prox, will not draw hundreds or thousands of amp but something much less. A shorted cell battery is not a short to ground of the entire battery, which of course would draw amps at a catastrophic rate.

So I am hoping that someone can show me some numbers

If you hook a 12 volt battery to a one dead cell 10 volt battery, what is the current draw. The potential difference between the 12 volt and 10 volt is two, there is still an internal resistance in the battery.

In our case, I assumed a dead short, we smelled some gas, and when checked, the batteries were hot. The furthest one was about a 3 second battery, ie I could hold my hand on it for 3 seconds, ( I know not very scientific) the others were progressively cooler.

IF a one cell short was a dead short to ground with 4400 ah available from the other battery, and if there were hundreds if not thousands of amps flowing as per your theory, I should have seen the insulation from the battery leads melting off the wires. There was not. And this is what makes me think that the amp draw into a one shorted cell battery is not in the thousands of amps.


If say the cell closest to ground on a single isolated 12 volt battery shorted, would your theory support the process that the remaining 5 cells would pump big amps through to ground.

 

[DCockey]

.....


If say the cell closest to ground on a single isolated 12 volt battery shorted, would your theory support the process that the remaining 5 cells would pump big amps through to ground.

The cells in a battery are in series. The current through each cell is the identical (unless there is a short to the exterior of the battery). It does not make any difference which battery is shorted.

MCA is the rating of how much current a fully charged battery at 32F/0C can supply at a minimum of 7.2 volts for at least of 30 seconds. Typical MCA ratings for marine batteries are 600 amps or greater. Assuming the current will be proportional to the voltage difference then the current flowing at a voltage of 10 volts at the battery would be at least several hundred amps.